rhoneyager/libicedb

Scattering element ordinal number

Closed this issue · 1 comments

Scattering element number is always zero.
Shouldn't it be a progressive number?

Right now, the code doesn't write the scattering element numbers when the numbers are trivial (1, 2, 3, 4, ...). Ordinarily, I would just consider it to be optional (and space-wasting) data. However, the scattering element number variable is also a dimension.

HDF5's Dimension Scale API suggests defining the variable but never writing data to it. The HDF5 developers have discussed adding support for sequential "fill values".

I'll fix it.